Mechanism analysis is an effective tool for evaluating the feasibility of complex mechanisms,
but in order to find the worst conditions (singular solution) which cause problems such as wearing,
there is a need for parameter combination analysis of huge volumes, causing the cost for
calculation and time required for analysis to increase exponentially.
This paper explores the usage of Efficient Global Optimisation (EGO) which uses Kriging
(Gaussian) response surface technique and couples it with mechanism analysis software allowing
to reduce the total runs needed while handling intractable design space which leads to simulation
failure. This technique is used for improving the design of steam turbine valve which regularly
suffers from surface wear.
